Dark Matter Credit: The Development of Peer-to-Peer Lending and Banking in France

157 Pages·2019·26.41 MB·English

by  Philip T. Hoffman, Gilles Postel-Vinay, and Jean-Laurent Rosenthal

Dark Matter Credit
THE PRINCETON ECONOMIC HISTORY OF THE
WESTERN WORLD
Joel Mokyr, Series Editor
THE DEVELOPMENT OF PEER-TO-PEER
LENDING AND BANKING IN FRANCE
A list of titles
in this series
appears at the back
of the book Philip T. Hoffman
Gilles Postel-Vinay
Jean-Laurent Rosenthal
PRINCETON UNIVERSITY PRESS
PRINCETON AND OXFORD
Contents
(cid:9) (cid:9)
Copyright © 2019 by Princeton University Press Acknowledgments vii
Published by Princeton University Press
Introduction 1
41 William Street, Princeton, New Jersey 08540
Chapter 1. 1740 and the Rules of the Game 10
6 Oxford Street, Woodstock, Oxfordshire OX20 1TR
press.princeton.edu Chapter 2. Spatial Variety versus Centralization:
Change in Eighteenth-Century Credit Markets 48
All Rights Reserved
Library of Congress Control Number 2018953892 Chapter 3. The Revolution: Collapse, Reform, and
Modeling the Space of Debt 74
ISBN 978-0-691-18217-9
British Library Cataloging-in-Publication Data is available Chapter 4. Networks of Knowledge 95
Editorial: Joe Jackson & Samantha Nader Chapter 5. The Brief but Significant Life of an
Institutional Innovation 122
Production Editorial: Ali Parrington
Text and Jacket/Cover Design: Leslie Flis Chapter 6. The Diffusion of Banks: Peer-to-Peer Credit
Markets as Substitutes for Banks 149
Production: Erin Suydam
Publicity: Tayler Lord Chapter 7. Banks and Notaries 176
This book has been composed in Sabon
Chapter 8. Prices Return 195
Printed on acid-free paper.-
Chapter 9. Conclusion 218
Printed in the United States of America
13579108642 Appendices 239
Notes 265
Bibliography 285
Index 299
LIB
332
.11
HOF
Acknowledgments
ti
The Russell Sage Foundation generously supported this project in its early
stages, and the California Institute of Technology, the Paris School of Eco-
nomics, and UCLA helped bring it to fruition. We would also like to
thank the Center of European and Eurasian Studies and the Collins Fund
at UCLA for financial support, and the École des Hautes Études en Sci-
ences Sociales and the Paris School of Economics for hosting Hoffman
and Rosenthal in Paris.
We have benefited enormously from the comments and suggestions
made by many colleagues: Jérôme Bourdieu, Tracy Dennison, Dave
Grether, Steve Haber, Lionel Kesztenbaum, Naomi Lamoreaux, Annie
Lunel, Thierry Magnac, Eric Monnet, Larry Neal, Oscar Gelderblom, Tim-
othy Guinnane, Joost Jonker, Steve Quinn, Paula Scott, William Sewell,
Akiko Suwa-Eisenmann, Peter Temin, Francesca Trivellato, Joachim
Voth, Jan de Vries, Kirstin Wandschneider; the referees at Explorations in
Economic History and Social Science History, where two related articles
were published; Karen S. Cook, Margaret Levi, and Russell Hardin, who
edited a volume in which another related article appeared; participants at
seminars at Berkeley, Harvard, the Paris School of Economics, the Russel
Sage Foundation, Yale, and USC; members of the audience in talks at
meetings of the All-UC Economic History group, the Economic History
Association, and the Social Science History Association; two readers at
Princeton University Press, who gave us valuable criticisms and advice
about improving our manuscript; our editor at Princeton, Joe Jackson,
who nudged us along with guidance and encouragement; and, last but
not least, Joel Mokyr, who has helped us and so many other economic
historians to write better economic history. This book would never have
been finished without the unstinting efforts of Laura Betancour, Maria
Chichtchenkova, Alena Lapatniovna, and Asli Sumer, who helped photo-
graph and code much of the data. We deeply appreciate their expert assis-
tance. Two economic historians inspired us, long ago, to work on debt: the
late Lance Davis and Maurice Lévy-Leboyer. They deserve our thanks.
Dark Matter Credit
Introduction
Anemic financial markets mire countries in poverty. There are other rea-
sons why countries remain poor, but a feeble financial system blocks
economic growth; or so both modern econometric evidence and histori-
cal studies seem to show. Banks and credit markets are particularly
important—even essential.' Without banks, incomes languish, but when
they open their doors, lending surges, and economic growth takes off.
This argument has become a commonplace. Yet it is hard to reconcile
with an inconvenient fact: that somehow much of Europe managed to
grow rich long before banks became widespread in the nineteenth cen-
tury.2 If the usual argument is correct, the wealthy parts of Europe should
have been penniless too, for, without banks, they—like the rest of Europe—
ought to have been condemned to poverty. But they were prosperous by
the standards of the day, not poor.
Could it be that credit abounded in Europe even before banks spread
across the continent? That was the question we set out to answer, using
data for France. Since France (unlike Italy, England, or the Low Coun-
tries) has long been considered a laggard in developing banks, it was an
ideal test case, because as early as the eighteenth century, much of the
country was clearly well off by world standards.' How, then, could it have
grown wealthy in the eighteenth century, and even richer in the nineteenth,
without having a large number of banks? Could the French tap other, hid-
den sources of credit and do so on a large scale? If so, then borrowers in
other leading countries could likely do the same.
As this book shows, there were ways to borrow in France before banks
opened their doors, and the mountain of debt this shadow credit system
raised was big, even by modern standards. As early as 1740, the system
allowed nearly a third of French families to borrow; if measured relative to
GDP, then by 1840 it was mobilizing as much credit for mortgages as the
United States' banking system did in the 1950s.4 Moreover, much of this
capital was raised for agriculture and urban real estate, sectors critical in
a developing economy that banks often shun because of the risks of
farming and the long loan maturities of real estate lending.
Until now, virtually no one has noticed this big debt, despite its size. In
a way, it is like the dark matter that makes up some eighty-five percent of
the universe but cannot be directly observed. And while astronomers and
physicists can infer the existence of dark matter from its effects, econo-
mists, historians, and other social scientists are not that lucky. Worse yet,
they have simply assumed that what cannot easily be observed—private
2 • Introduction(cid:9) Introduction . 3
credit in the past or in poor countries today-was insignificant or simply Table 1. Estimates of notarized lending in France
not there at all.'
Year
That assumption is mistaken, as is the argument that banks are an
essential first step toward mobilizing large amounts of financial capital 1740 1780 1807 1840 1865 1899
and building a thriving debt market for private borrowers. And that is
Number of loans in year 437 368 362 556 395 265
not all that is wrong either. France, we found out, eventually got more
(thousands)
banks than anyone imagined. If these banks were a more efficient source
of credit, as the claim about their importance supposes, then their prolif- Number of outstanding 1,696 1,477 856 1,419 1,328 1,645
loans (thousands)
eration should have made the shadow lending disappear. But it did not
vanish. Indeed, it persisted in France, and elsewhere too, up to World Value of loans 161 336 329 772 914 1180
War I, and was only killed off by government intervention that tipped the (million livres/francs)
scales in favor of banks. The reason was that banking and the shadow Stock of outstanding debt 1426 2398 1120 3650 4150 7690
lending system were not competing sources of credit. Rather, they comple- (million livres/francs)
mented one another, so that both thrived together.
We know all this because we actually measured the dark matter of Maturity (years, 5.8 4.3 2.4 2.6 3.4 6.2
private credit before 1900, rather than just supposing it was trivial (see unweighted)
table 1). We also counted the number of banks using new historical evi- Maturity (years weighted 8.9 7.1 3.4 4.7 4.5 6.5
dence. Private credit, we learned, was big and pervasive, and not at all by loan value)
challenged by the diffusion of banks in the nineteenth century. If anything,
Per capita stock of debt 58.0 86.9 37.7 104.6 109.0 191.5
our measurements are likely underestimates, because they omit lending
(livres/francs)
that we did not count even though it might be substantial.'
We reached these conclusions for France thanks to unique fiscal records Stock of debt to GDP 15.8 22.8 9.6 27.2 19.9 23.6
that survive for the period 1740-1931. These records let us gather the nec- (percent)
~
essary data at relatively low cost. We thought it would be worth exploit-
Source: Estimates from our sample. For details, see chapter 1.
ing them because of the large amount of lending we had already uncovered
Note: For GDP estimates for France after 1800, we relied on Toutain (1987). Because
in Paris using a different source of evidence.? It was not at all clear, how- there are no GDP estimates for France before 1800, we simply assumed total income was
ever, that the example of Paris would generalize, for two reasons. First, growing at 0.4 percent per year from 1740 to 1780, and again from 1780 to 1807.
Paris had an unusually large number of wealthy investors who could fund Netting out population growth leads per capita income to grow at 0.1 percent per year
before 1780, and 0.125 percent from 1780 to 1807. Monetary amounts in 1740 and 1780
loans. Furthermore, the city's lenders, borrowers, and potential financial
are in livres, the money of account before the French Revolution; for 1807-99, they are in
intermediaries dwelled near one another and might interact repeatedly,
francs, the currency created during the French Revolution. For the years of our cross
which would make it easier to arrange loans. Conditions would not be sections, they both had the same value in silver.
the same elsewhere, particularly where credit markets were thin and where
lenders, borrowers, and intermediaries lived too far apart even to find one
another. The question was whether Paris was atypical, and the fiscal records it raises other important questions. To begin with, how was credit allocated
gave us the answer. before banks? The big debt, it turns out, consisted of thousands of bilateral
Those records are peculiar to France, but the evidence they yield can loans, loans that matched up a borrower and a lender, as in modern peer-to-
be compared with data from Germany, Great Britain, and the United peer lending. These loans were sizable, had maturities that were frequently
States. The comparison shows that France is not at all unusual. The shadow two years or more (see table 1), and often involved people who did not
credit system flourished in the past in these other wealthy countries too, know each other. For such loans, lenders cannot simply assume borrowers
and it may loom large in many developing economies as well, if research- will repay, and charging a higher interest rate to offset the risk may attract
ers take the time to measure it. nothing but deadbeats who have no intention of paying off their debts.
Our discovery of all the debt financed by the shadow credit system not Securing the loans with collateral may not solve the problem, either. How
only overturns the standard argument about banks and economic growth; does a lender tell what a pledged property is worth and how that value will
4 • Introduction Introduction • 5
evolve, particularly in an economic downturn like that which struck the poster child for the economic damage caused by barriers to the develop-
US mortgage market after 2006? Borrowers usually have a good sense ment of a banking system. Economic historians have long believed that a
both of their creditworthiness and of the value of their collateral, but delayed spread of banks in France retarded the country's industrializa-
lenders' information is typically much skimpier. In the language of eco- tion and slowed economic growth. Economists have pushed the argument
nomics, lenders' and borrowers' information is asymmetric. further, blaming the French legal system for hindering financial develop-
Unlike some peer-to-peer lending on the web today, the bilateral loans ment, not just in France itself, but in all the countries around the world
in the past were arranged by a network of brokers. The brokers not only that inherited its particular brand of civil law. Those two claims turn out
brought the borrowers and the lenders together, but overcame the prob- to be wrong too. Nothing blocked bank entry in France, and that is why
lems of asymmetric information, which afflict all credit markets. That was we found that the country in fact had far more banks than economic
true not just in Paris, where the brokers interacted with one another historians thought. French civil law did not hamper financial innovation
repeatedly in a way that could easily spread information about creditwor- either. More generally, while France may not have been the leading econ-
thiness; it was also true in small towns, where their dealings would be omy in Europe, its performance was good enough to provide resources for
much rarer. Even there our brokers certified borrowers and their collateral, three centuries of military competition, first with the vast and rich Haps-
and gave lenders better information. That proved essential to building a burg empires and later with England.
large stock of debt to GDP. Our discoveries have significant implications for the world today. To
The brokers, both in Paris and the rest of France, were notaries, gov- begin with, they cast doubt on the evidence backing the claim that ane-
ernment sanctioned keepers of legal records in countries influenced by mic conventional financial markets have impeded economic growth in
Roman law, who combined the preservation of records with the roles of poor countries. The claim is supported by cross-country regressions, but
lawyer, financial adviser, and real estate broker. Their network arose the regressions assume that private lending outside of banks and other
because the records they kept revealed what collateral was worth and who modern financial intermediaries is measured accurately. If this sort of pri-
was a good credit risk. The information they could cull from their records vate lending is not measured accurately, then the true relationship between
allowed the notaries to match up lenders with creditworthy borrowers and financial development and economic growth—so our French evidence
so solve debt markets' vexing informational problems. suggests—may well be far weaker than everyone assumes.lo
The solution therefore grew out of a peculiar feature of Roman law. Successfully measuring private credit has other significant implications.
That itself is a surprise, for Roman law, and its modern offspring—the In particular, it corrects the standard story of how credit markets develop.
civil law that holds sway in continental Europe and Latin America—are That story begins in a world of no lending and then traces a small set of
thought to hobble financial development.$ Yet in France, as we shall innovations (such as stock and bond exchanges or big universal banks
see, this infrastructure of Roman law nurtured a thriving financial struc- with branches and a variety of services). It focuses on these innovations
ture. The structure did evolve in a different direction from its British because they spread internationally, as people learned how to imitate the
counterpart, which may in fact have been biased toward banks. Both financial innovators and how to copy their institutions and organiza-
financial systems, however, did fund economic growth, and by 1900, Paris tions.11 Yet change in credit markets has never followed this sort of unique
was, like London, an international financial center. The two financial sys- path, and neither has financial development more generally, either in the
tems had started apart and followed dissimilar paths as they developed, past or in poor countries today. Financial development, it turns out, can
but by 1913 they both had large thriving equity and debt markets. 9 take many different routes to abundant credit and easy mobilization of
That is not all we uncovered. We also analyze how lending in the financial capital, and the road selected depends on politics, on inequality,
shadow credit system was shaped by geography and the growth of cities. on economic shocks and legal institutions, and on the spatial develop-
Since cities had more savers with large sums to lend, borrowing in a city ment of cities and the economy. No one has analyzed this long-run process
might be appealing, but the cost of travel ruled out long trips to find a loan. of change until now. We do in this book, which reaches back over two
We work out how the network of notaries dealt with travel costs and centuries and continues through industrialization and across enormous
urban savings, and we chart how their dealings changed over time. political and social upheavals, ranging from the French Revolution and
Finally, beyond simply assessing how the shadow credit system was the Napoleonic Empire to the rise of democracy and World War I.
affected by the diffusion of banks, we also determine whether any obstacles Along the way, we learn how private credit markets in France func-
slowed bank entry—an important topic since France has been held up as a tioned in the past and how they changed as the economy grew, partly as
6 • Introduction(cid:9) Introduction • 7
a result of shifts in demand, and partly as a result of shifts in supply, driven borrowing—an impact that was negative in the short run but overwhelm-
by institutional innovations and political and legal innovations. We see ingly positive over the long run.1 2 These enduring economic consequences
how borrowers and lenders devised new loan contracts, created ingenious of the French Revolution have long been neglected, particularly the long-
ways of securing loans, and made the transition from ancient ways of lend- run positive ones, which had echoes outside of France.
ing (annuities and medium-term loans with a balloon payment) to the Outside of history, sociologists will benefit from the questions we raise
modern mortgage. We also find out how financial capital was mobilized about the common method of analyzing networks that are limited to sim-
across space in the era before railroads, when transportation was rudi- ilar individuals. So will economists who work on networks. Similarly, legal
mentary. And, above all else, we discover how our brokers solved the scholars and political scientists will profit from the doubt we cast on the
daunting problems of asymmetric information in credit markets, and did widespread argument that civil law condemns an economy to economic
so on a large scale, long before the arrival of modern banks and stock stagnation. The same goes for political scientists who believe that politi-
exchanges and the creation of government lien registries and private credit cal institutions shape economic development.
ratings. Our conclusions are derived from the French data, but they are To make all these discoveries, we had to proceed differently from econ-
likely to apply to credit in other economies as well, because in most parts omists or historians who study credit markets. Unlike economists who
of Western Europe borrowers and lenders could avail themselves of very have focused heavily on the recent experience of developing countries, we
similar sets of contracts and information systems. reach back and study credit in a diverse set of localities over nearly two
Figuring out how these credit markets worked required more than mea- centuries. And, unlike historians, we have not done a local study. Instead,
surement alone. We also had to model how borrowers, lenders, and bro- we gathered extensive quantitative data and estimated medium and long-
kers acted. The economic models, which are explained in plain language term private indebtedness for the economy as a whole. We needed all this
for readers unfamiliar with economics, proved essential. They made our argu- data to analyze the network of brokers and to gauge the impact of banks
ments precise, let us test our claims, and revealed what was happening as more and more of them opened their doors. The data had to extend
when the historical sources fell silent. Without them, we would still be back in time well before the Industrial Revolution and stretch forward
trying to make sense of all the dark matter of private credit. through the nineteenth century as banks proliferated and the economy
The story we tell about the evolution of private credit will interest not developed. And it had to continue into the twentieth century to see what
just readers in economics, but in history, law, and in all the social sciences. finally killed off the shadow credit system.
Historians, for instance, will gain a new perspective on the social and eco- We begin our book by describing the data that revealed how much pri-
nomic history of lending. The large historiography devoted to the subject vate credit there was and how loans were arranged. The bulk of this evi-
of credit has invoked debt to explain both peasant immiseration and the dence concerns 239,269 individual loans and the variables that affected
expansion of markets, and assumes personal ties between debtors and lending in a sample of ninety-nine French credit markets. The markets
creditors to characterize a noncapitalist economy. Much of this literature, ranged from Paris to small villages, and for each market, we gathered the
though, is limited to a particular locality, using local account books, family data for six years (1740, 1780, 1807, 1840, 1865, and 1899). For a subset
papers, or loan contracts that have survived in one particular place. Much of these years, we also gathered evidence from seventy-three additional
of it is confined to traditional periods of historical study as well—in France, markets. Beyond these two large samples, we collected much smaller sam-
the Old Regime, or the French Revolution and the Napoleonic Empire, ples in 1912, 1927, and 1931 to chart the demise of peer-to-peer lending.
or the century from 1815 to 1914. In this book we broke free of these So that readers can understand how we measured private debt, we
restrictions, because we want to chart the evolution of credit across nearly explain the construction of our samples and the legal and political insti-
two centuries of massive legal and organizational change, including the tutions that governed the credit market. We then estimate the size of the
coming of banks. And we wanted to measure lending for the whole econ- market in 1740 and explore who was involved in it (chapter 1). The next
omy, not one particular locality, and see how it changed over time and how issue is determining what boosted the volume of private lending between
different credit markets were related. 1740 and 1780 (chapter 2). Prominent among the explanations were inno-
By using our evidence as a benchmark, historians who undertake new vative loan contracts and better ways of protecting lenders against default.
local studies of credit can now ask how lending in their locality was con- The background in these first two chapters is essential for another reason
nected with other markets. Historians will also be able to assess, for the as well: it lays out the problems private credit markets faced and how
first time, the lasting impact that the French Revolution had on private these peer-to-peer lending systems operated.
8 • Introduction Introduction • 9
Grasping how private credit markets function also requires assessing the truth. A notary, it turns out, had nothing to fear from bankers, who
the impact the French Revolution had on lending institutions. The private dared not compete with the notary in his own specialty of mortgage lend-
credit market was laid low by hyperinflation during the French Revolu- ing (unless, of course, they had government backing and a government
tion, but in the long run it benefited from the revolution's institutional monopoly, like the Cr6dit Foncier). The bankers and notaries in fact
reforms, such as the creation of lien registries, which helped protect lend- focused on different corners of the credit market, and their businesses were
ers. Although these reforms took decades to diffuse, they helped the credit complementary: they reinforced one another.
market recover completely from the damage done by the revolutionary Surprisingly, the huge number of loans that we discovered in the dark
inflation. After assessing the effect of the inflation (in chapter 3), we explore matter credit market were almost all made at one interest rate: one price.
these new institutions and then analyze how notaries matched up lenders This outcome—a priceless equilibrium in the language of economics—
with creditworthy borrowers. When a notary could not find a match derived from usury legislation and from the incentives created by the
among his own clients, he referred the prospective borrower or lender to asymmetric information in the private credit markets. Prices only began
other nearby notaries, whom he cooperated with in what would become to matter again (they had played a role in French private credit markets in
a local lending network. The resulting networks linked markets through- the seventeenth century) in the late nineteenth and early twentieth century,
out France and overcame local imbalances of supply and demand. when the government began to intervene in the market on a large scale.
How all this happened only became clear when we built our economic As we show in chapter 8, the government first provided financial backing
models in chapters 3 and 4. In the process, we analyzed how the notaries to a large mortgage loan bank, the Cr6dit Foncier, and gave it a monopoly
made referrals and what that implied for the spatial distribution of on the issuance of mortgage-backed securities. Then the government
borrower-lender matches. It was impossible to test the models against evi- started subsidizing loans to private borrowers.
dence from the notaries' business records, which do not survive. But we This first history of dark matter credit markets carries important les-
could test them against data from the fiscal records. Remarkably, the fis- sons for financial markets and governments today, as we suggest in the
cal data support our models and reject a very different interpretation of conclusion (chapter 9). One lesson is that there is no single path to finan-
the notaries' behavior. cial development. Another is that existing traditional financial institutions
We also investigated how the notaries interacted with other financial may be far more important than anyone supposes. Replacing them may
intermediaries, such as banks—the subject of chapters 5 through 7. The therefore be a mistake and may leave new market entrants (such as mod-
notaries were innovative, and in the nineteenth century they devised a new ern banks) vulnerable to problems of adverse selection when they get stuck
type of loan contract that involved dealing with bankers and merchants, with all the bad credit risks. Finally, a third lesson is that banks are not
as we show in chapter 5. This new contract and earlier innovations by likely to enter mortgage markets unless they have government backing.
notaries both run counter to the claim that countries such as France would Otherwise, even the largest banks run the risk of falling victim to default-
be slow to develop financially, because they were governed by the sup- ing borrowers, as happened in the 2008 financial crisis. All three lessons
posedly rigid Napoleonic civil law. In reality, civil law was far more flex- should not be forgotten.
ible than many scholars believe, and it certainly did not keep notaries from
discovering new ways of doing things.
To measure the interaction between banks and notarial credit, we gath-
ered new data on the number of banks in France in the nineteenth and
early twentieth centuries. Chapter 6 analyzes the spread of banks in France
and compares their diffusion with similar data for the United Kingdom.
France had more banks than anyone imagined, and it erected no barriers
to bank entry. If France did end up with proportionally fewer banks than
England, it was because of demand and—surprisingly—because of the
relative weakness of the British peer-to-peer credit system.
Chapter 7 then asks whether banks were so much more efficient that
they drove notaries out of the business of arranging peer-to-peer loans
as they spread across France. As we discovered, nothing was further from
'11
1740 and the Rules of the Game • 11
CHAPTER 1
1740 and the Rules of the Game
Rouen
A* • •
•
• • q Paris
The big debt we discovered consisted of peer-to-peer loans, long before • Pontivy
•¤
that term emerged on the web. In France there were millions of them, even y •
• r
centuries ago. What were they like? Here is one example: in 1740 Jean
Pajot traveled eleven kilometers from his home to the town of Bellac • lie
in central France (see figure 1.1) to borrow forty livres (about two or Bellac (cid:9) ® • ¤ •
q Paris • •
three months pay for a rural laborer) from Guillaume Reymond.t Pajot was • •• •
not alone, even in Bellac. Other borrowers from the town and its environs © Rouen and Lyon •i ¤ p•L yon
had local notaries draw up over one hundred loan contracts that year, ¤ Markets pop> 10000 • • ~ •¤
totaling twenty thousand livres. • •
• Markets pop< 10000 ~
Since Bellac and the nearby villages in this remote part of France ••(cid:9) •
counted only some 8,500 inhabitants in 1740, it might seem, at least at • •~ •• o • •
first glance, that relatively few people were taking out peer-to-peer loans. •• ¤ •~ •
•
But if we consider households rather than individuals, the participation
rate was far from trivial. If each household averaged four persons, then
seven percent of local households took out loans in 1740. And since loans
typically had to be paid back in two years, some fourteen percent of house- Figure 1.1. The sample.
holds would owe money in this market at any time. The number of lend- Source: See text.
ers would be smaller, because many lenders made multiple loans, but it
still seems likely that at least twenty percent of the households in Bellac
were involved in notarial credit in 1740, either as borrowers or lenders. of the major components of the national debt—the government's General
That is a significant fraction. Farm (Ferme Générale), which was only half the size .2 When averaged over
And Bellac is only one example, for borrowers were taking out similar the whole population, outstanding notarial debt came to almost sixty livres
numbers of loans across France. If we take all of the ninety-nine markets per person in 1740, or two months of per capita income. Even more sur-
in our sample together and extrapolate to France as a whole, then at least prising, the 1.7 million loans outstanding suggest that notarial lending
430,000 loans were made in 1740, for a total of 160 million livres, and allowed nearly one third of France's six million households to borrow, even
some 1.7 million debt contracts were outstanding, worth 1.4 billion livres more than what we found for Bellac.
(table 1.1). This mountain of private, nongovernmental debt raises some serious
These numbers are large. The stock of notarial debt, even though it questions. First of all, how can we reconstruct past lending in a society,
excluded nearly all commercial and consumer credit, amounted to sixteen particularly for credit markets that have long been shrouded from view?
percent of GDP in 1740 (table 1). Although that may at first glance seem How do we know what happened, and how reliable is our knowledge?
paltry, especially when compared to the level of mortgage debt accumu- Second, even if we can accurately count these loans, add up their values,
lated in some economies on the eve of the 2008 crisis, it is more than what or average their maturities, how can we speak of credit markets back in
mortgage markets achieve in many developing economies today. And while 1740? After all, this is a half century before the French Revolution, in what
it totaled somewhat less than what the government owed its creditors historians call Old Regime France, and it is not exactly the obvious place to
(some two billion livres in 1740), it was still huge. The volume of lending look for capital markets. Like most of Western Europe back then, it was an
coursing through the notarial credit market every year in fact dwarfed one absolutist monarchy, with an economy as yet unsullied by industrialization.